[rrd-users] Re: Spikes (1.0.8)

[Tobias Oetiker]

Yesterday you sent me mail regarding [rrd-users] Re: Spikes (1.0.8):

*> I am seeing the same sort of thing.  It's not just limited to 1.0.8;  I've
*> seen it in 1.0.7 also.  Here's a more complete description:
*> 
*> It's a 4 timescale graph.  1 pdp/row, 6 pdp/row, 24 pdp/row,  288 pdp/row.
*> 
*> Data starts accumulating.   The spike appears in the second time scale.  It
*> appears at a point *for which no data has been recorded yet*.  (As in, the
*> set has collected 150 samples, and the spike appears at scale 2 about 1000
*> samples out.)  It is also far larger or far smaller than any of the samples
*> which have been recorded at the previous time scale.  In some data sets the
*> spike is negative when all the data recorded in the first time scale is
*> positive, and it's usually several magnitudes larger/smaller than any of the
*> data points in the first scale.

it would be great if I could get some code demonstrating this behaviour ...
it would make debugging so much simpler ... 

what happens when you look at the data with dump? Is there realy a spike in
the data? What happens to the spike whan you continue to use the system?
does it get overwritten eventually as it lies in the future?

Anybody else seeing data inconsistancies in places where no data should be
goping anyway ?

cheers
tobi
*> 
*> - Jeff Younker - jeff at mdli.com - These are my opinions, not MDL's -
*> 
*> --
*> * To unsubscribe from the rrd-users mailing list, send a message with the
*>   subject: unsubscribe to rrd-users-request at list.ee.ethz.ch
*> 
*> 

-- 
 ______    __   _
/_  __/_  / /  (_) Oetiker, Timelord & SysMgr @ EE-Dept ETH-Zurich
 / // _ \/ _ \/ / TEL: +41(0)1-6325286  FAX:...1517  ICQ: 10419518 
/_/ \.__/_.__/_/ oetiker at ee.ethz.ch http://ee-staff.ethz.ch/~oetiker

--
* To unsubscribe from the rrd-users mailing list, send a message with the
  subject: unsubscribe to rrd-users-request at list.ee.ethz.ch